NC.7.SP.1 · NC.7.SP.2Print or save as PDF
Random Sampling
How could a school learn how many students ride the bus without asking every single one? It can ask a small group and use the answers to make a smart guess. In this guide you will learn how to pick a fair group. You will see why a fair group gives a good guess and an unfair group does not. Then you will take many samples and see how much the guesses change.
Source: Authored for NCMath (not from the PDF prototype) — needs teacher review
Samples and populations
Words to know
Pollsters, scientists, and stores all use samples to learn about big groups. Here are the words you will see in this guide.
| Word | What it means |
|---|---|
| Population | The whole group you want to know about. |
| Sample | A smaller part of the population that you actually look at. |
| Representative sample | A sample that looks like the population. Its parts match the whole group. |
| Random sample | A sample picked by chance. Every member of the population has an equal chance to be picked. |
| Biased sample | A sample that leans one way. Some members of the population are more likely to be picked than others. |
| Estimate | A good guess about the population, made from a sample. |
| Inference | A conclusion about a population that you make from sample data. |
| Variation | How much the estimates change from one sample to the next. |
| Simulation | A model that uses random numbers to act out a chance event, so you do not have to try it for real. |
Why we use a sample
A school has 200 sixth graders. The office wants to know how many ride the bus. Asking all 200 students takes a long time.
So the office asks a small group and uses the answers to learn about everyone. That small group is a sample. The whole group of 200 is the population.
In the picture, each dot is a student. A filled dot means the student rides the bus. In real life, nobody has this picture. That is why we need a sample.
Example
Ms. Lee asks 20 of the 200 sixth graders if they ride the bus. What is the population? What is the sample?
The population is the whole group she wants to know about. That is all 200 sixth graders. The sample is the small group she asks. That is the 20 students. Check: every student in the sample is also in the population, and the sample is smaller.
The population is the whole group she wants to know about. That is all 200 sixth graders. The sample is the small group she asks. That is the 20 students. Check: every student in the sample is also in the population, and the sample is smaller.
Representative samples
When an inference is valid A conclusion about a population is valid only if the sample is representative of that population. — representative sample → valid inference
Suppose 40 out of every 100 students in a school ride the bus. A good sample would show about 40 riders out of every 100 students in it. That sample looks like the whole school. We call it representative.
You can only trust what a sample tells you about the population if the sample is representative. If the sample does not look like the population, the guess will be off.
Example
You want to know the favorite sport of the 480 students at Park Middle. Sample A is 40 students from the basketball team. Sample B is 40 names drawn from a hat holding all 480 names. Which sample is representative?
Sample B is representative. Every student could be picked, so it should look like the whole school. Sample A is not. Basketball players are more likely to pick basketball than other students are. Check: Sample A leaves out most of the school.
Sample B is representative. Every student could be picked, so it should look like the whole school. Sample A is not. Basketball players are more likely to pick basketball than other students are. Check: Sample A leaves out most of the school.
Trap: A common mistake is thinking a bigger sample is always a good sample. If you ask 100 basketball players, you still only hear from basketball players. A big sample that is not representative still gives a bad guess.
Random samples
How do you get a representative sample? Pick by chance, not by choice. In a random sample, every member of the population has an equal chance to be picked.
You can put every name in a hat and draw 20. You can also give each student a number and let a computer pick 20 numbers.
In the picture, the orange rings show 20 students picked at random. Seven of them ride the bus. That is 7 out of 20, or 35%. The whole school is 40%. A random sample is usually close, but it is not perfect.
Example
A teacher writes each of her 30 students' names on a slip of paper. She mixes the slips and draws 6. Is this a random sample? What if she picks the first 6 students who sit down?
Drawing slips is random. Every student has the same chance, 6 out of 30, to be picked. The first 6 who sit down are not random. Students who come early are different from students who come late. Check: in a random sample, nobody chooses who gets in.
Drawing slips is random. Every student has the same chance, 6 out of 30, to be picked. The first 6 who sit down are not random. Students who come early are different from students who come late. Check: in a random sample, nobody chooses who gets in.
Biased samples
Now suppose you stand at the bus loop and ask 20 students there. Every one of them rides the bus. You would think 100% of the students ride the bus. That is not true. The sample leans toward bus riders.
A sample like this is biased. Some students were much more likely to be picked than others.
Here is how to tell the two kinds of samples apart.
| How the sample is picked | Random or biased? | Why |
|---|---|---|
| Draw names from a hat with every student | Random | Everyone has an equal chance. |
| Ask students at the bus loop | Biased | Bus riders are more likely to be asked. |
| Ask only your friends | Biased | Your friends are like you. Other students are left out. |
| Let anyone click a poll on a website | Biased | Only people who care enough to click are in the sample. |
Example
Mr. Ortiz asks 20 students at the bus loop. 18 of them ride the bus. He says that 90% of the school rides the bus. Is that a valid inference?
18 out of 20 is 90% of the sample. But the sample was picked at the bus loop, so it is biased. It is not representative. The inference is not valid. We cannot say 90% of the school rides the bus. Check: the true value for this school is 40%, far from 90%.
18 out of 20 is 90% of the sample. But the sample was picked at the bus loop, so it is biased. It is not representative. The inference is not valid. We cannot say 90% of the school rides the bus. Check: the true value for this school is 40%, far from 90%.
Trap: A common mistake is thinking that picking people by hand is fair if you try to be fair. Picking easy people, such as friends, feels fair. But it is still biased. Only chance makes a sample random.
From a sample to the population
Back to the random sample of 20 students. Seven ride the bus. That is 7 out of 20, or 35%.
If the sample is representative, the whole school should look about the same. So about 35% of 200 students ride the bus. The double number line shows that 7 of 20 matches 70 of 200.
This is an estimate. It is a good guess, but it is not exact.
Example
A random sample of 25 of the 500 members of a gym is taken. 10 of the 25 prefer to work out in the morning. Estimate how many of the 500 members prefer mornings.
10 out of 25 is 40%. 40% of 500 is 200. About 200 members prefer mornings. Check: 500 is 20 times as many as 25, and 10 × 20 = 200.
10 out of 25 is 40%. 40% of 500 is 200. About 200 members prefer mornings. Check: 500 is 20 times as many as 25, and 10 × 20 = 200.
Trap: A common mistake is thinking an estimate is exactly right. Our sample gave 70 bus riders. The picture shows 80 riders. The estimate is close, but it is not the same.
Many samples show how much estimates vary
One sample is not the whole story
If you take a different random sample, you get a different estimate. That is variation.
Here, 30 different random samples of 20 students were taken from the school. Each dot is the percent who ride the bus in one sample.
The dots run from 15% to 60%. But they pile up near 40%. The center of the dots is the best guess for the whole school.
Example
How many of the 30 samples gave an estimate from 30% to 50%? What does the plot say about the whole school?
Count the dots at 30, 35, 40, 45, and 50. There are 2 + 5 + 5 + 3 + 7 = 22 dots. So 22 of the 30 samples are in that range. The dots pile up near 40%. So I infer that about 40% of the school rides the bus. Check: 22 out of 30 is a big majority, and the real school has 80 out of 200, which is 40%.
Count the dots at 30, 35, 40, 45, and 50. There are 2 + 5 + 5 + 3 + 7 = 22 dots. So 22 of the 30 samples are in that range. The dots pile up near 40%. So I infer that about 40% of the school rides the bus. Check: 22 out of 30 is a big majority, and the real school has 80 out of 200, which is 40%.
Trap: A common mistake is trusting one sample too much. One random sample gave 15%. Another gave 60%. Both were fair samples. That is why we look at many samples.
Bigger samples vary less
The plot shows two sets of samples from the same school. In the top set, each sample has 10 students. In the bottom set, each sample has 40 students.
The dots for samples of 10 are spread out. The dots for samples of 40 are bunched together. A bigger sample gives estimates that are closer to each other, and closer to the truth.
Example
Find how wide each set of dots is. The samples of 10 run from 0% to 80%. The samples of 40 run from 32.5% to 52.5%. How do the widths compare?
For samples of 10, 80 − 0 = 80 percentage points. For samples of 40, 52.5 − 32.5 = 20 percentage points. The bigger samples are 4 times narrower. Check: 80 ÷ 20 = 4.
For samples of 10, 80 − 0 = 80 percentage points. For samples of 40, 52.5 − 32.5 = 20 percentage points. The bigger samples are 4 times narrower. Check: 80 ÷ 20 = 4.
Trap: A common mistake is thinking a bigger sample makes the estimate exactly right. Even the samples of 40 ranged from 32.5% to 52.5%. They vary less, but they still vary.
Finding an unknown percent
A jar holds 500 beads. Some are red. Nobody knows how many. A student takes a random sample of 20 beads, records the percent that are red, and puts the beads back. Then the jar is shaken and she does it again. She takes 25 samples in all.
The dots run from 15% to 70%, but most are near 35%. The center is about 35%. So I infer that about 35% of the beads are red.
Because the samples vary, the true percent could be a bit higher or lower. Most dots are between 25% and 45%. So the true percent is probably in that range.
Example
Use the plot to estimate how many of the 500 beads are red. Give a range that is likely to hold the true number.
The center is about 35%. 35% of 500 is 175. So about 175 beads are red. Most samples are between 25% and 45%. 25% of 500 is 125 and 45% of 500 is 225. The true number is probably between 125 and 225. Check: when the jar was counted, 180 beads were red. That is 36%, and it fits in the range.
The center is about 35%. 35% of 500 is 175. So about 175 beads are red. Most samples are between 25% and 45%. 25% of 500 is 125 and 45% of 500 is 225. The true number is probably between 125 and 225. Check: when the jar was counted, 180 beads were red. That is 36%, and it fits in the range.
Simulated samples
Sometimes it is hard to try something many times for real. A simulation uses random numbers to act it out.
A basketball player makes 70% of her free throws. Random digits from 0 to 9 can stand for her shots. Seven of the ten digits are 0 through 6. So the digits 0 to 6 mean made, and 7 to 9 mean missed.
One game is 10 shots, so we use 10 digits and count the made shots. The plot shows 30 simulated games.
- Decide what the digits mean. Digits 0 to 6 are made shots and 7 to 9 are misses.
- Make 10 random digits for one game.
- Count the made shots in the game.
- Repeat this for many games, 30 here.
- Show the counts on a dot plot and look at the pattern.
Example
In the 30 simulated games, how many games had 4 or fewer made shots? What percent of the games is that?
Count the dots at 4. There are 3 dots, and none below 4. So 3 of the 30 games had 4 or fewer made shots. 3 ÷ 30 = 0.10, which is 10%. A 70% shooter makes 4 or fewer in about 1 game out of 10. Check: most dots are near 7, which is 70% of 10 shots.
Count the dots at 4. There are 3 dots, and none below 4. So 3 of the 30 games had 4 or fewer made shots. 3 ÷ 30 = 0.10, which is 10%. A 70% shooter makes 4 or fewer in about 1 game out of 10. Check: most dots are near 7, which is 70% of 10 shots.
Quick guide to common mistakes
| If you need to... | Do this | Watch out for |
|---|---|---|
| Pick a sample | Pick by chance, such as names from a hat or random numbers | Do not pick friends or easy people. That is biased. |
| Decide if an inference is valid | Ask if the sample looks like the whole population | A big sample that is biased is still not valid. |
| Estimate a number for the population | Find the percent in the sample, then take that percent of the population | The estimate is close, but it is not exact. |
| Use many samples | Look at where the dots pile up | Do not trust one sample alone. |
| Pick a sample size | Bigger samples vary less | Even big samples vary a little. |
| Predict with a simulation | Repeat many times and count how often an outcome happens | A few trials are not enough. |
Use this table as a last check before you hand in your work.
Memory hook: Pick by chance. Pick enough. Look at many samples.
Practice
1. Principal Diaz wants to know if the 480 students at Park Middle like the new bell schedule. She asks 40 of them.
(a) What is the population?
(b) What is the sample?
2. Two students want to know the favorite kind of music of all the students at their school.
(a) Ava draws 30 names from a hat that holds every student's name.
(b) Ben asks 30 students at the school's band concert.
Which sample is random? Which is biased?
3. Jo asks 20 students in the band room if music is their favorite class. 14 of them say yes. Jo says, "70% of the students in our school pick music as their favorite class." Is Jo's inference valid? Explain.
4. A random sample of 40 of the 800 students at a school is asked how they get to school. 14 of the 40 walk. Estimate how many of the 800 students walk.
5. A bag holds 400 tiles. Some are green. Maria takes a random sample of 20 tiles and records the percent that are green. She puts them back, mixes, and repeats this 20 times. The dot plot shows her results.
(a) About where is the center of the dots?
(b) About how many of the 400 tiles are green?
6. Plots A and B show percents from random samples of the same school. Each plot has 25 samples. One plot used samples of 10 students. The other used samples of 40 students.
(a) Which plot used samples of 40?
(b) If you can take only one sample, which size should you use? Why?
7. Principal Reyes takes three different random samples of 50 students from her school of 600 students. She asks if they like soccer. Sample 1 has 21 yes answers. Sample 2 has 18. Sample 3 has 24.
(a) Find the percent of yes answers in each sample.
(b) Use the lowest and highest percents to estimate the least and most students in the school who like soccer.
8. A goalie saves 60% of penalty shots. To simulate, use random digits from 0 to 9. The digits 0 to 5 mean a save and 6 to 9 mean a goal. One game is 10 digits (10 shots). The dot plot shows the number of saves in 25 simulated games.
(a) How many of the 25 games had 3 or fewer saves?
(b) What percent of the games is that?
(c) A real goalie who saves 60% of shots saves only 3 in one game. Is that very unusual?